3.1.1. Methodology

We calculated a basic community characteristic, the specific richness, to see if patterns could be detected in the study area. The same calculation as for Chapter 1 have been performed for the considered stations.

Assumption: A higher species richness indicates a more desirable status.

3.1.2. Application

The following map represent the corresponding EQR status for each station:

3.2.1. Methodology

We calculated basic community characteristics, the total density and biomass of individuals, to see if patterns could be detected in the study area. The same calculations as for Chapter 1 has been performed (with the addition of biomass data) for the considered stations.

3.2.2. Application

3.3.1. Methodology

We calculated basic community characteristics, the Shannon diversity and Pielou evenness, to see if patterns could be detected in the study area. The same calculations as for Chapter 1 has been performed for the considered stations.

Assumption: A higher diversity indicates a more desirable status.

3.3.2. Application

The following maps represent the corresponding EQR status for each station:

3.4.1. Methodology

We calculated a basic community characteristic, the taxonomic distinctness, to see if patterns could be detected in the study area. The same calculations as for Chapter 1 has been performed for the considered stations.

3.4.2. Application

3.5.1. Methodology

We studied functional diversity based on these species traits:

• body composition (non calcified tissue, calcareous, calcareous calcium carbonate, calcareous amorphous calcium carbonate, calcareous aragonite, calcareous calcite, calcareous high magnesium calcite, chitinous)
• body size (small, medium, large)
• food diet (filter, surface deposit, subsurface deposit, predator, scavenger, grazer, parasite)
• mobility (sessile, limited, mobile)
• lifestyle (fixed, tubicolous, burrower, crawler, swimmer)

Species were assigned to a trait using a binary code (0: absence of the trait, 1: presence). This allowed to calculate functional richness, divergence and evenness according to Laliberté & Legendre (2010).

3.5.2. Application

3.6.1. Methodology

AMBI is an ecological index that is used to detect a perturbation in an ecosystem based on its communities (Borja et al., 2000). This perturbation is linked with the organic matter loading, according to Pearson and Rosenberg (1978)’s model.

To compute this index, species are classed into five groups in relation to their tolerance to this perturbation:

• group I (GI): vulnerable species
• group II (GII): indifferent species
• group III (GIII): tolerant species
• group IV (GIV): first-order opportunistics
• group V (GV): second-order opportunistics

These groups are based on expert opinion on the physiology of species and experimental studies, but the attribution of a species to a group can be somewhat arbitrary (e.g. based on related phyla information) so it needs to be interpretated carefully. The AMBI index (also called biotic coefficient) is continuous between 0 to 6, and is calculated using this formula:

\[ AMBI = \frac{\sum_{i}^{GI-V} w_{i} . P_{i}}{100} \]

• \(P_{i}\) is the proportion of each group (percentage of the total number of species)
• \(w_{i}\) is the weighting parameter of each group (respectively 0, 1.5, 3, 4.5 and 6)
• \(i\) is the ecological group

3.6.2. Application

3.7.1. Methodology

M-AMBI is a complementary method that is used to calculate an Ecological Quality Ratio (EQR), a measure of the good environmental status. It is based on a multivariate ordination of the stations using the AMBI index, the species richness and the Shannon diversity. The result givises a value between 0 and 1 after comparison to reference values.

Assumption: A higher richness and diversity indicates a more desirable status.

These values are called “references” but this needs to be discussed as this vision is limited. They have been set with the 95 % percentile of the distribution. This is a recommendation by Nicolas Desroy, so that we do not detect an increase of EQR when there is a small perturbation (see work by Pearson & Rosenberg and the Intermediate Disturbance Hypothesis).

This calculation yielded 21 for S and 2.53 for H.

3.7.2. Application

The following map represent the corresponding EQR status for each station.

No clear tendancy can be discovered here, apart from the fact that the overall status seems to be “High”. Several hypothesises can explain this result:

• the M-AMBI index describes reality well, so that overall perturbation from organic matter is low
• there is a bias in the index due to the species classification in groups, originally suited for European ecosystems
• the assumptions for the reference values are not correct
• the configuration of the bay makes the perturbation small relative to the water volume and bathymetric condition

Further work is needed to determine the individual responses of somes species, along with the use of different methods to understand other perturbations and cumulative impacts.

3.8.1. Methodology

BOPA is an index that uses a relative abundance ratio of species in a community to infer a state of perturbation. Ratios with many species have been tested, and opportunistic polychaetes and amphipods have been selected to be the most pertinent (originally to detect effects of an oil-spill on soft-bottom communities, e.g. from the Sea Empress or the Amoco Cadiz). It indicates an absence of pollution when amphipods are dominant and a pollution when opportunistic polychaetes are dominant. It has been updated from its original form in 2000, and varies between 0 and \(log_{10}(2)\) (~ 0.3).

The equation is:

\[ BOPA = \left( \frac{f_{P}}{f_{A} + 1} + 1 \right) \]

• \(f_{P}\) is the relative frequency of opportunistic polychaetes (abundance / total density)
• \(f_{A}\) is the relative frequency of amphipods (abundance / total density)

We considered AMBI groups GIII to GV for polychaetes and GI for amphipods (without Jassa genera).

3.8.2. Application

These are the polychaetes and amphipods present in our species list (including the confidence score used during group classification).

The following map represent the corresponding EQR status for each station. To use EQR classification, we used the conversion method from Dauvin & Ruellet (2007).